# Uncovering Critical Features for Deepfake Detection through the Lottery Ticket Hypothesis

arXiv: https://arxiv.org/abs/2507.15636 · HTML: https://mlsanity.com/paper/2507.15636/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 19 of its 29 references that are in the map.
- First posted on arXiv 21 Jul 2025.

## Builds on (most cited first, 19 of 19)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks](https://mlsanity.com/paper/1803.03635/) (2018)
- [FaceForensics++: Learning to Detect Manipulated Facial Images](https://mlsanity.com/paper/1901.08971/) (2019)
- [SNIP: Single-shot Network Pruning based on Connection Sensitivity](https://mlsanity.com/paper/1810.02340/) (2018)
- [LLM-Pruner: On the Structural Pruning of Large Language Models](https://mlsanity.com/paper/2305.11627/) (2023)
- [Celeb-DF: A Large-scale Challenging Dataset for DeepFake Forensics](https://mlsanity.com/paper/1909.12962/) (2019)
- [The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scale](https://mlsanity.com/paper/1811.00982/) (2018)
- [Picking Winning Tickets Before Training by Preserving Gradient Flow](https://mlsanity.com/paper/2002.07376/) (2020)
- [Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes](https://mlsanity.com/paper/2101.10994/) (2021)
- [MesoNet: a Compact Facial Video Forgery Detection Network](https://mlsanity.com/paper/1809.00888/) (2018)
- [What Do Compressed Deep Neural Networks Forget?](https://mlsanity.com/paper/1911.05248/) (2019)
- [FaceForensics: A Large-scale Video Dataset for Forgery Detection in Human Faces](https://mlsanity.com/paper/1803.09179/) (2018)
- [Coarsening the Granularity: Towards Structurally Sparse Lottery Tickets](https://mlsanity.com/paper/2202.04736/) (2022)
- [Pruning for Robust Concept Erasing in Diffusion Models](https://mlsanity.com/paper/2405.16534/) (2024)
- [Deepfakes, Misinformation, and Disinformation in the Era of Frontier AI, Generative AI, and Large AI Models](https://mlsanity.com/paper/2311.17394/) (2023)
- [RTMobile: Beyond Real-Time Mobile Acceleration of RNNs for Speech Recognition](https://mlsanity.com/paper/2002.11474/) (2020)
- [3D Point Cloud Network Pruning: When Some Weights Do not Matter](https://mlsanity.com/paper/2408.14601/) (2024)
- [Adaptive Knowledge Distillation for Classification of Hand Images using Explainable Vision Transformers](https://mlsanity.com/paper/2408.10503/) (2024)

## Built on by (most cited first, 3 of 3)

- [Reliable Audio Deepfake Detection in Variable Conditions via Quantum-Kernel SVMs](https://mlsanity.com/paper/2512.18797/) (2025)
- [Quantum Kernels for Audio Deepfake Detection Using Spectrogram Patch Features](https://mlsanity.com/paper/2605.06035/) (2026)
- [On Evaluating Quantum Kernel Robustness for Low-Resource Cross-Corpus Audio Deepfake Detection](https://mlsanity.com/paper/2610.00649/) (2026)

## Abstract (arXiv)

Recent advances in deepfake technology have created increasingly convincing synthetic media that poses significant challenges to information integrity and social trust. While current detection methods show promise, their underlying mechanisms remain poorly understood, and the large sizes of their models make them challenging to deploy in resource-limited environments. This study investigates the application of the Lottery Ticket Hypothesis (LTH) to deepfake detection, aiming to identify the key features crucial for recognizing deepfakes. We examine how neural networks can be efficiently pruned while maintaining high detection accuracy. Through extensive experiments with MesoNet, CNN-5, and ResNet-18 architectures on the OpenForensic and FaceForensics++ datasets, we find that deepfake detection networks contain winning tickets, i.e., subnetworks, that preserve performance even at substantial sparsity levels. Our results indicate that MesoNet retains 56.2% accuracy at 80% sparsity on the OpenForensic dataset, with only 3,000 parameters, which is about 90% of its baseline accuracy (62.6%). The results also show that our proposed LTH-based iterative magnitude pruning approach consistently outperforms one-shot pruning methods. Using Grad-CAM visualization, we analyze how pruned networks maintain their focus on critical facial regions for deepfake detection. Additionally, we demonstrate the transferability of winning tickets across datasets, suggesting potential for efficient, deployable deepfake detection systems.